Katherine, 

Do you have the answer to your past question ? 

I have the same question. I think that the parametric modulators extracted in SPM.xX.X should be set as conditions and not as a modulators. But I don't know how.

If someone has an answer, it would be appreciated 

Isabelle

Originally posted by Katherine Swett:

Hi Alfonso! I hope you have been doing well (and staying healthy!). I am running a parametric modulation analysis in conn, using the recommendations in this thread. I have one question: it seems that in order to enter the first-level covariate files, other users have pulled the SPM.xX.X vector, which I believe is the hrf-convolved covariate. I was wondering whether putting an hrf convolved time series in the set-up is acceptable, or whether this results in two rounds of convolution that might skew the data? I appreciate your help! Best, Katherine Aboud Originally posted by Alfonso Nieto-Castanon:

Dear Ewa, Yes, in the latest releases of CONN you may actually perform those PPI-style analyses in CONN directly. To do that you would need to create a series of first-level covariate files (.txt or .mat files) containing your modulator variables and enter those as additional first-level covariates in Setup.CovariatesFirstLevel. Then, in the first-level anlaysis tab you may select the analysis type 'other temporal modulation' effects, and select there each of these covariates, and that will perform a PPI analyses looking at the connectivity modulation associated with your time or distance-to-target variables (i.e. those areas where functional connectivity strength covaries with time or distance-to-target). Unfortunately the 'import' functionality will not automatically read the information of your modulatory variables from your SPM.mat files so you would instead need to manually/progammatically create those first-level covariate files for each subject/session (I will try to add the functionality to automatically import into CONN that info from SPM-defined task-modulation variables to the next release). Best Alfonso   Originally posted by Ewa Miendlarzewska:

Dear Alfonso, I am uncertain if this comment also applies to v15.a of conn. I have a 1st level design in SPM with a time modulator (ordinal values 1-n of trials) and a modulator by distance to target (continuous values 0-1). Unfortunately, they don't show when I load my SPMs to conn. Under covariate, conn recognizes only the motion parameters which it calls "SPM covariates". Am I doing something wrong? Thank you for your help. Best, E